1. Field of the Invention
The present invention relates to an image processor, an image processing method, and an endoscope system that calculates data, such as numerical values provided for diagnosis using an endoscopic image, using the endoscopic image captured by an endoscope.
2. Description of the Related Art
In the medical field, diagnosis using endoscope systems including a light source device, an endoscope, and a processor device has been performed widely. In the diagnosis using the endoscope systems, an insertion part of the endoscope is inserted into a subject, illumination light is radiated from a distal end part, and an observation object (such as a mucous membrane within the subject) irradiated with the illumination light is imaged by an imaging sensor loaded on the distal end part of the endoscope. An image of the observation object is generated using image signals obtained by the imaging, and is displayed on a monitor.
In the endoscope systems, generally, an image (hereinafter referred to as an observation image) in which the observation object is observable in natural tone is displayed by radiating white illumination light (also referred to as normal light) to image the observation object. Moreover, endoscope systems that obtain an image (hereinafter referred to as a special observation image) in which blood vessels, pit patterns, and the like of the observation object are enhanced by using light having a specific wavelength band as the illumination light have also spread. Since information on the blood vessels, the pit patterns, and the like are important diagnosis material, the special observation image in which these are enhanced is particularly useful for diagnosis.
Additionally, in recent years, endoscope systems or diagnosis support devices that support doctors' diagnosis by calculating the depth, thickness, and density of blood vessels, and diagnosis indexes obtained from color balance of pixels representing the blood vessels, using the endoscopic image (the normal observation image, the special observation image, or image signals for generating these) captured by the endoscope are also known (JP4493637B and JP5395725B (corresponding to U.S. Pat. Pub. No. 2011/245,642)).
Information on blood vessels (hereinafter referred to as blood vessel information) that can be calculated using the endoscopic image as in JP4493637B and JP5395725B is information useful for diagnosis as described above. However, doctors do not perform diagnosis on the basis of one kind of blood vessel information out of the blood vessel information, such as the depth, thickness, and density of the blood vessels, and the diagnosis indexes obtained from the color balance of pixels representing the blood vessels, but complexly perform diagnosis in consideration of a plurality of items of blood vessel information, including blood vessel information that is not evaluated by related-art endoscope systems other than the above blood vessel information. For example, the thickness of the blood vessels and the density of the blood vessels are respectively blood vessel information useful for diagnosis. However, the state of the observation object is not necessarily differentiated only by the thickness of the blood vessels being a specific thickness, or the density of the blood vessel being a specific density. Diagnoses is performed from multifaceted and complex points of sight by taking into consideration a plurality of items of blood vessel information that the state of the observation object is a specific lesion because the thickness of the blood vessels is equal to or greater than the specific thickness and the density of the blood vessels is equal to or greater than beyond the specific value.
In accordance with the actual conditions of the multifaceted and complex diagnosis as described above, in recent years, it is required that endoscope systems or image processors for analyzing an endoscopic image calculate information or the like that is more direct and useful than the blood vessel information calculated in JP4493637B and JP5395725B, and support doctors' diagnosis.